Positron emission tomography (PET) devices and single photon emission computed tomography (SPECT) devices are used to capture an image representing the internal function of a subject. In the photographing by a PET device, a drug labeled with various positron emissions is administered to the subject. The PET device detects the radiation caused by that positrons emitted from the various positron emissions combine with negatrons and disappear. In the photographing by an SPECT device, a drug labeled with a radioactive isotope is administered to the subject. The SPECT device detects the radiation emitted from the radioactive isotope. Besides, X-ray computed tomography (CT) devices are used to capture an image representing the internal structure of a subject. An X-ray CT device irradiates the subject with X-rays from different directions a plurality of times, and detects X-rays having passed through the subject with an X-ray detector.
These devices generate an image representing the internal structure of the subject based on projection data obtained by the detection. At this time, some devices apply a reconstruction process to the projection data by a block iterative successive approximation method to generate the image. In the block iterative successive approximation method, the projection data is divided into a plurality of subsets, and successive approximation is performed on each of the subsets. When such division of the projection data into subsets (subsetting) is performed, the time required for the successive approximation (the time required for the reconstruction process) is reduced compared to the case without the subsetting.
Meanwhile, the subsetting causes degradation of the image quality. This is due to a bias in statistical errors contained in each subset. In other words, the block iterative successive approximation is known as a method in which an increase in the number of subsets reduces the time required for the reconstruction process and also degrades the image quality. Accordingly, the generation of an image with good image quality requires less number of subsets and long-time reconstruction process. If the number of subsets is increased to reduce the time taken for the reconstruction process, the image quality is lowered.